An industrial robot end displacement high-precision detection method based on photoelectric interference

Abstract

The application discloses a kind of industrial manipulator end displacement high-precision detection methods based on photoelectric interference, it is related to the field of precision measurement and detection, comprising: S1: acquisition photoelectric interference signal and down frequency filtering to obtain complex envelope signal;S2: based on complex envelope signal cumulative phase, calculate preliminary displacement observation value;S3: acquisition equipment information constructs sensing dataset, generates multi-source error state vector;S4: utilize extended Kalman filter to remove error to generate purified displacement value;S5: according to real-time environmental parameter, calculate air refractive index, and combine reference air refractive index to compensate purified displacement value, generate correction displacement value. Through multi-source sensor fusion, Kalman filter error estimation and air refractive index compensation, realize the high-precision, stability and robustness measurement of industrial manipulator end displacement.

Description

Technical Field

[0001] This invention relates to the field of precision measurement and detection technology, specifically a high-precision method for detecting the displacement of the end effector of an industrial robotic arm based on photoelectric interferometry. Background Technology

[0002] Industrial robotic arms are widely used in high-precision assembly, micro-manipulation, and manufacturing tasks in complex environments. Displacement detection technology is a crucial element for the motion control and operational accuracy of the robotic arm's end effector. Existing displacement detection methods mainly include encoder-based joint angle detection, vision-based spatial position measurement, and high-precision detection based on laser interferometry. Among these, the encoder method is limited by the accumulation of errors in the joint transmission chain, making it difficult to guarantee the accuracy of the absolute displacement at the end effector; vision-based measurement has poor robustness under varying lighting conditions, occlusion, and complex environments, and its accuracy is also limited by imaging resolution and calibration errors; laser interferometry has extremely high resolution and stability, but in complex industrial environments, it still faces interference from factors such as fluctuations in interference signal quality, changes in environmental refractive index, thermal effects, and structural vibrations, leading to certain deviations in the detection results.

[0003] Therefore, how to improve the stability and accuracy of displacement detection at the end of a robotic arm in complex industrial environments has become a pressing technical problem to be solved in this field. Summary of the Invention

[0004] Based on the shortcomings of the prior art described above, the purpose of this invention is to provide a high-precision detection method for the end-effector displacement of an industrial robotic arm based on photoelectric interferometry, so as to solve the above-mentioned technical problems.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a high-precision detection method for the end-effector displacement of an industrial robotic arm based on photoelectric interferometry, comprising:

[0006] S1: Acquire the two orthogonal interference electrical signals output by the photoelectric interference module, perform digital down-conversion and filtering processing, and obtain the complex envelope signal;

[0007] S2: Based on the complex envelope signal, differential operation is performed to obtain the instantaneous phase increment, which is then accumulated to obtain the accumulated optical phase. According to the mapping relationship between the accumulated optical phase and the laser wavelength, the preliminary displacement observation value is calculated, and the displacement observation base set is constructed.

[0008] S3: Collect device information to construct a sensor dataset and generate a multi-source error state vector;

[0009] S4: Based on the extended Kalman filter algorithm, the multi-source error state vector is optimally estimated, and the optimally estimated multi-source error state vector is separated from the preliminary displacement observations to generate the purified displacement value;

[0010] S5: Calculate the air refractive index based on the real-time collected environmental parameters, and compensate the purification displacement value by combining it with the preset reference air refractive index, and finally generate the corrected displacement value.

[0011] The present invention is further configured such that S1 includes:

[0012] A photoelectric interference module is installed at the end of the robotic arm to emit a measurement beam that is directed at a fixed reference mirror. The measurement beam is a laser with a fixed wavelength.

[0013] The interference light signal is received by a photodetector and converted into two orthogonal interference voltage signals. The two synchronously sampled orthogonal interference voltage signals are then combined into a complex measurement signal.

[0014] The complex measurement signal is multiplied by a preset digital local oscillator sequence to perform digital down-conversion, wherein the angular frequency of the digital local oscillator sequence is equal to the frequency shift angular frequency of the acousto-optic modulator in the photoelectric interference module;

[0015] The down-converted signal is subjected to digital low-pass filtering to obtain a complex envelope signal.

[0016] The present invention is further configured such that S2 includes:

[0017] The complex envelope signal sequence is first subjected to short-time complex domain smoothing to suppress high-frequency noise, and then the difference between adjacent sampling points is calculated to construct a complex difference.

[0018] The instantaneous phase increment is obtained by taking the imaginary part of the ratio between the complex difference signal and the complex envelope signal, based on the first-order approximation principle.

[0019] The instantaneous phase increment is accumulated over time to form an accumulated optical phase based on the initialization time. The initial value of the complex envelope signal is sampled and set by the system under static conditions, and the accumulated optical phase is set to zero during phase accumulation initialization.

[0020] The present invention is further configured to monitor the complex envelope signal in real time and extract the magnitude amplitude of the complex envelope signal as a quality indicator;

[0021] When the quality index is lower than the preset threshold, the signal quality is determined to be faulty, the phase accumulation process is paused and the resampling process is triggered.

[0022] The present invention is further configured such that when the quality index is not lower than a preset threshold, the cumulative optical phase is multiplied by a preset proportional coefficient to obtain a preliminary displacement observation value, wherein the proportional coefficient is the ratio of the laser vacuum wavelength to a constant four times pi, and the laser vacuum wavelength is taken from the laser parameters of the photoelectric interference module.

[0023] Add the current system timestamp and quality index to each preliminary displacement observation;

[0024] By integrating accumulated optical phase, preliminary displacement observations, timestamps, and quality indicators, a basic set of displacement observations is constructed.

[0025] The present invention is further configured such that S3 includes:

[0026] The device collects multi-source sensor data to construct a sensor dataset, which includes: joint position number, joint position coordinates, joint angle, joint and structural temperature, triaxial accelerometer signal, and gyroscope signal.

[0027] The theoretical end position is calculated based on the joint position number, joint position coordinates, joint angle, preliminary displacement observation, and preset nominal kinematic parameters. The theoretical end position is then compared with the preliminary displacement observation to obtain the residual signal.

[0028] Based on the residual signal, the correction amount related to the nominal kinematic parameters is estimated and mapped to the kinematic bias.

[0029] Based on the temperature of each joint and structure of the robotic arm and the working time of the actuator, a thermal state data sequence is constructed. Based on the thermal state data sequence and the preset thermal expansion model, the displacement change caused by the thermal expansion of the structure is calculated, and the displacement change is set as the thermally induced displacement deviation.

[0030] The present invention is further configured to construct a vibration state data sequence based on the triaxial accelerometer signal and gyroscope signal at the end of the robotic arm, calculate the deviation of the end vibration position based on the vibration state data sequence and the structural dynamics model, and set the deviation as the vibration displacement deviation.

[0031] Based on the time series of vibration displacement deviation at the end of the robotic arm, the vibration velocity change at consecutive moments is calculated, or the vibration velocity information is directly extracted using gyroscope signals, and the calculation result is used as the vibration velocity term.

[0032] A multi-source error state vector is generated by combining kinematic deviation, thermal displacement deviation, vibration displacement deviation, and vibration velocity term.

[0033] The present invention is further configured such that S4 includes:

[0034] A state transition model and an observation model are established based on the multi-source error state vector, and the extended Kalman filter method is used to estimate and update the multi-source error state vector in real time.

[0035] The Kalman-filtered and updated multi-source error state vector is removed from the initial displacement observations to obtain the purified displacement value.

[0036] The present invention is further configured such that S5 includes:

[0037] Real-time environmental parameters are collected by barometric pressure sensors, temperature sensors, and humidity sensors to construct environmental status data;

[0038] Environmental state data is input into a preset refractive index estimation model, which outputs the current air refractive index. The refractive index estimation model is constructed based on a polynomial approximation method.

[0039] The present invention is further configured to call a reference air refractive index preset during the system calibration stage, compare the current air refractive index with the reference air refractive index, and obtain the air refractive index deviation;

[0040] The air refractive index deviation is input to the displacement compensation unit, which maps the refractive index deviation into a displacement correction amount based on a preset calibration coefficient.

[0041] Based on the purification displacement value, the displacement correction value is subtracted from the purification displacement value to finally generate the correction displacement value.

[0042] This invention provides a high-precision method for detecting the end-effector displacement of an industrial robotic arm based on photoelectric interferometry. The method comprises the following steps: S1: Acquiring two orthogonal interference electrical signals output from a photoelectric interferometry module, performing digital down-conversion and filtering to obtain a complex envelope signal; S2: Performing differential operations on the complex envelope signal to obtain the instantaneous phase increment, accumulating it to obtain the accumulated optical phase, and calculating the preliminary displacement observation value based on the mapping relationship between the accumulated optical phase and the laser wavelength, thus constructing a basic displacement observation set; S3: Acquiring equipment information to construct a sensor dataset and generating a multi-source error state vector; S4: Optimizing the multi-source error state vector using an extended Kalman filter algorithm, separating the optimally estimated multi-source error state vector from the preliminary displacement observation value, and generating a purified displacement value; S5: Calculating the air refractive index based on real-time acquired environmental parameters, and compensating the purified displacement value using a preset reference air refractive index, ultimately generating a corrected displacement value. The beneficial effects include:

[0043] By introducing multi-source sensor data and establishing a joint error model of kinematics, thermal expansion, and vibration, and combining extended Kalman filtering to achieve dynamic estimation and real-time compensation, multiple interference factors can be effectively eliminated, significantly improving the robustness and accuracy of displacement detection.

[0044] By constructing an air refractive index estimation model based on polynomial approximation and combining it with real-time environmental parameters to compensate for optical measurement errors, the stability and reliability of displacement detection under different temperature, humidity and pressure environments are ensured.

[0045] By digital down-conversion and phase accumulation processing of the interferometric signal, combined with quality index monitoring and resampling mechanisms, continuous, stable and high-resolution displacement observation results can be obtained, effectively avoiding error accumulation caused by signal failure and phase jump.

[0046] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0048] Figure 1 The flowchart illustrates a high-precision detection method for the end-effector displacement of an industrial robotic arm based on photoelectric interferometry, as an exemplary embodiment of the present invention. Detailed Implementation

[0049] The embodiments of the present invention will be described below with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention and not for limiting the scope of protection of the present invention.

[0050] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0051] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0052] Example:

[0053] A high-precision method for detecting the end-effector displacement of an industrial robotic arm based on photoelectric interferometry, such as... Figure 1 As shown, it includes:

[0054] S1: Acquire the two orthogonal interference electrical signals output by the photoelectric interference module, perform digital down-conversion and filtering processing, and obtain the complex envelope signal;

[0055] S2: Based on the complex envelope signal, differential operation is performed to obtain the instantaneous phase increment, which is then accumulated to obtain the accumulated optical phase. According to the mapping relationship between the accumulated optical phase and the laser wavelength, the preliminary displacement observation value is calculated, and the displacement observation base set is constructed.

[0056] S3: Collect device information to construct a sensor dataset and generate a multi-source error state vector;

[0057] S4: Based on the extended Kalman filter algorithm, the multi-source error state vector is optimally estimated, and the optimally estimated multi-source error state vector is separated from the preliminary displacement observations to generate the purified displacement value;

[0058] S5: Calculate the air refractive index based on the real-time collected environmental parameters, and compensate the purification displacement value by combining it with the preset reference air refractive index, and finally generate the corrected displacement value.

[0059] The present invention is further configured such that S1 includes:

[0060] A photoelectric interference module is installed at the end of the robotic arm to emit a measurement beam that is directed at a fixed reference mirror. The measurement beam is a laser with a fixed wavelength.

[0061] The interference light signal is received by a photodetector and converted into two orthogonal interference voltage signals. The two synchronously sampled orthogonal interference voltage signals are then combined into a complex measurement signal.

[0062] The complex measurement signal is multiplied by a preset digital local oscillator sequence to perform digital down-conversion, wherein the angular frequency of the digital local oscillator sequence is equal to the frequency shift angular frequency of the acousto-optic modulator in the photoelectric interference module;

[0063] The down-converted signal is digitally low-pass filtered to obtain a complex envelope signal. Specifically, a photoelectric interferometer probe mounted at the end of the robotic arm emits a fixed-wavelength laser beam towards a fixed reference mirror. The reference beam and the measurement beam interfere at the reference mirror, and the interference intensity is split into two approximately orthogonal photoelectric components by a 90-degree optical mixer. The two photoelectric components are converted into voltage signals by a high-speed photodetector, and after filtering and amplification by the analog front-end circuit, they enter a two-channel synchronous analog-to-digital converter and are acquired at the same sampling time. The two synchronous digital samples are synthesized into complex measurement samples in a predefined order within the FPGA for processing in the complex domain. The digital local oscillator module inside the FPGA generates a corresponding complex local oscillator sequence according to the frequency shift frequency of the acousto-optic modulator; subsequently, the system performs complex domain multiplication on each complex measurement sample and its corresponding complex local oscillator sample in a streaming manner, down-converts the high-frequency carrier component, and shifts the required phase envelope to the baseband frequency band. The down-converted complex signal passes through an anti-aliasing low-pass filter implemented within the FPGA, filtering out high-frequency components and retaining the in-phase and quadrature baseband components that change slowly with the end displacement. Each output sample obtained after filtering is the complex baseband envelope signal, with its real and imaginary parts corresponding to the in-phase and quadrature samples required for subsequent phase processing, respectively. During the output stage, the system performs real-time quality judgment on the envelope amplitude. If the amplitude meets the quality threshold, the complex envelope samples are packaged frame by frame, timestamped, and marked with a quality identifier, and transmitted to the embedded host via DMA or a real-time bus. If the amplitude is below the threshold, a failure is marked in the data frame, and a resampling or alarm process can be triggered. The entire processing chain is implemented in a pipelined manner on the FPGA / hardware accelerator, ensuring deterministic latency and high throughput. In the physical implementation, amplifier gain calibration, ADC bias calibration, and local oscillator frequency / phase calibration must be completed during system startup to ensure that the amplitude and phase of the complex envelope signal are consistent with subsequent phase statistical processing.

[0064] The present invention is further configured such that S2 includes:

[0065] The complex envelope signal sequence is first subjected to short-time complex domain smoothing to suppress high-frequency noise, and then the difference between adjacent sampling points is calculated to construct a complex difference.

[0066] The instantaneous phase increment is obtained by taking the imaginary part of the ratio between the complex difference signal and the complex envelope signal, based on the first-order approximation principle.

[0067] The instantaneous phase increment is accumulated over time to form an accumulated optical phase based on the initialization time. The initial value of the complex envelope signal is sampled and set by the system under static conditions, and the accumulated optical phase is set to zero during phase accumulation initialization.

[0068] Real-time monitoring of complex envelope signals, and extraction of the magnitude and amplitude of complex envelope signals as quality indicators;

[0069] When the quality index is lower than the preset threshold, the signal quality is determined to be faulty, the phase accumulation process is paused and the resampling process is triggered;

[0070] When the quality index is not lower than the preset threshold, the cumulative optical phase is multiplied by the preset proportional coefficient to obtain the preliminary displacement observation value. The proportional coefficient is the ratio of the laser vacuum wavelength to a constant four times pi. The laser vacuum wavelength is taken from the laser parameters of the photoelectric interference module.

[0071] Add the current system timestamp and quality index to each preliminary displacement observation;

[0072] The system integrates accumulated optical phase, preliminary displacement observations, timestamps, and quality indicators to construct a basic set of displacement observations. Specifically, when the robotic arm is stationary, an interference signal under no-motion conditions is acquired through the photoelectric interferometry module as a reference envelope. The system records the complex initial value of this envelope and clears the accumulated phase register. During initialization, the amplitude threshold is adaptively calibrated to determine the threshold for subsequent quality judgment. After entering real-time processing mode, the S1 module continuously outputs the complex baseband envelope data stream. The data first enters the short-time complex domain smoothing module: using a calibrated short-window linear-phase FIR filter or a weighted moving average algorithm, the real and imaginary parts of the complex envelope signal are smoothed simultaneously. To avoid the group delay introduced by filtering affecting time synchronization, the filtering delay is recorded and compensated for in real time at the output. The processing result not only reduces high-frequency random noise but also improves the signal-to-noise ratio of the complex envelope, providing a stable input for subsequent differential operations. After the smoothing module outputs, it is fed into the difference module: based on the smoothed sample points cached by the system at the previous moment, a complex domain subtraction is performed with the current sample point to form a complex difference vector; if the magnitude of the current or previous smoothed sample point is lower than the initial gate threshold, the difference is marked as invalid, and subsequent steps skip this moment to avoid instantaneous phase estimation errors caused by low signal-to-noise ratio. For valid differences, the magnitude of the complex envelope denominator is first checked. If it is too small, a robust division strategy is automatically switched, such as adding protection against extremely small positive numbers or using historical sample backoff; under the condition of ensuring numerical stability, complex domain division is performed to calculate the complex ratio between the difference and the current complex envelope; the imaginary part is extracted from the complex ratio result as an approximation of the instantaneous phase increment. This step is implemented based on the first-order approximation principle and can complete the estimation of the instantaneous phase change without explicit formulas. For each estimated phase increment, the quality index is checked again by comparing it with a preset threshold and checking for abrupt amplitude changes. Only increments that pass the check are sent to the accumulation unit, and status information is recorded in the invalid interval. If failure persists, an automatic resampling or interpolation recovery strategy is triggered. The accumulation unit maintains a high-precision accumulated phase register and periodically performs temperature drift and numerical drift correction to ensure long-term accumulation accuracy. The high-precision accumulated phase is then converted into preliminary displacement observations by fixing the interferometric laser wavelength parameters during system calibration based on the preset mapping coefficients in the system configuration. Each preliminary displacement observation is then appended with the current system timestamp, complex envelope mode length amplitude, quality identifier, and filter delay compensation information. The above information is integrated to form a "displacement observation base set" containing accumulated phase, preliminary displacement observations, timestamps, and quality indexes, which can be directly called by the subsequent multi-source error fusion module. All processing steps employ a pipelined structure at the hardware level to ensure that complex envelope data enters the next step in real time; at the software level, an adaptive parameter interface is provided, which can automatically adjust parameters such as the filter window and gating threshold during operation to adapt to different environmental noise levels; a diagnostic and calibration interface is provided for maintenance and rapid recovery during long-term operation.

[0073] The present invention is further configured such that S3 includes:

[0074] The device collects multi-source sensor data to construct a sensor dataset, which includes: joint position number, joint position coordinates, joint angle, joint and structural temperature, triaxial accelerometer signal, and gyroscope signal.

[0075] The theoretical end position is calculated based on the joint position number, joint position coordinates, joint angle, preliminary displacement observation, and preset nominal kinematic parameters. The theoretical end position is then compared with the preliminary displacement observation to obtain the residual signal.

[0076] Based on the residual signal, the correction amount related to the nominal kinematic parameters is estimated and mapped to the kinematic bias.

[0077] Based on the temperature of each joint and structure of the robotic arm and the working time of the actuator, a thermal state data sequence is constructed. Based on the thermal state data sequence and the preset thermal expansion model, the displacement change caused by the thermal expansion of the structure is calculated, and the displacement change is set as the thermally induced displacement deviation.

[0078] Based on the triaxial accelerometer and gyroscope signals at the end of the robotic arm, a vibration state data sequence is constructed. Based on the vibration state data sequence and the structural dynamics model, the deviation of the end vibration position is calculated, and the deviation is set as the vibration displacement deviation.

[0079] Based on the time series of vibration displacement deviation at the end of the robotic arm, the vibration velocity change at consecutive moments is calculated, or the vibration velocity information is directly extracted using gyroscope signals, and the calculation result is used as the vibration velocity term.

[0080] A multi-source error state vector is generated by combining kinematic deviation, thermal displacement deviation, vibration displacement deviation, and vibration velocity term. Specifically, when the robotic arm enters the measurement mode, the control system synchronously collects the joint angle, end-effector acceleration, gyroscope angular velocity, and structural temperature of each joint via a real-time bus, and records the actuator running time. First, based on the collected joint angles and preset nominal kinematic parameters, the theoretical spatial coordinates of the robotic arm's end effector are calculated using a forward kinematics algorithm. Subsequently, these theoretical coordinates are compared with the preliminary displacement observations provided by the photoelectric interferometry module at each time step to obtain a residual signal representing the difference between the actual and theoretical end-effector positions. Using this residual signal, a recursive least squares method is executed within a sliding time window to estimate the correction amount related to the nominal kinematic parameters and map it to the kinematic deviation of the end effector space. Simultaneously, the system reads the real-time temperature data of each joint and structure, constructs a thermal state data sequence by combining the actuator's operating time information, and calls the multivariate linear thermal expansion model established during the calibration phase to calculate the structural dimensional changes caused by the temperature difference between the current environment and the calibration environment. This change is then mapped to the end effector spatial position to obtain the thermal displacement deviation. Subsequently, the end effector triaxial accelerometer and gyroscope continuously output dynamic signals. After bandpass and Kalman filtering, the data is input into the structural dynamics model. The end effector vibration displacement deviation is estimated by matching the modal response, and the vibration velocity change is calculated based on the time series of the vibration displacement deviation, or the end effector vibration velocity is directly derived from the gyroscope angular velocity, thus obtaining the vibration velocity term. Finally, the system sequentially integrates the kinematic deviation, thermally induced displacement deviation, vibration displacement deviation, and vibration velocity term into a timestamped multi-source error state vector, along with various quality indicators, for subsequent real-time error compensation and optimal estimation using the extended Kalman filter.

[0081] The present invention is further configured such that S4 includes:

[0082] A state transition model and an observation model are established based on the multi-source error state vector, and the extended Kalman filter method is used to estimate and update the multi-source error state vector in real time.

[0083] The cleaned displacement value is obtained by removing the Kalman-filtered and updated multi-source error state vector from the initial displacement observations. Specifically, the system receives the initialization information of the multi-source error state vector from S3 and the initial displacement observations provided by S2. The filter loads the pre-configured state definition, model type, and noise baseline, and completes the startup initialization at the rest time. During online operation, the filter performs a two-stage loop according to the sampling period: prediction and update. In the prediction stage, the state estimate from the previous time step is advanced to the current time step according to the model, accompanied by covariance prediction; after the observation arrives, the system performs time alignment and quality assessment on the observations, low-quality observations are weighted or removed, and high-quality observations enter the update process. In the update stage, the fusion weight is calculated by linearizing the observation function and the prediction is corrected by the observations, thereby obtaining the posterior state estimate and posterior uncertainty. The filter continuously monitors the residual statistics and adaptively adjusts the process or observation noise when necessary, or performs covariance correction to maintain numerical stability. After each update, the estimated error components are extracted from the posterior state and aligned with the corresponding initial displacement observation time, and these estimation errors are removed to obtain the cleaned displacement value. The final purified displacement value, along with the estimated uncertainty, timestamp, and quality identifier, is sent to the controller or upper-level fusion module for real-time closed-loop use. If the filter detects an anomaly or divergence trend, a predefined fault handling strategy is triggered, including resetting the state covariance, reverting to the most recent stable state, or entering a safe mode and reporting for maintenance.

[0084] The present invention is further configured such that S5 includes:

[0085] Real-time environmental parameters are collected by barometric pressure sensors, temperature sensors, and humidity sensors to construct environmental status data;

[0086] Environmental state data is input into a preset refractive index estimation model, and the current air refractive index is output. The refractive index estimation model is constructed based on a polynomial approximation method.

[0087] The reference air refractive index preset during the system calibration phase is called, and the current air refractive index is compared with the reference air refractive index to obtain the air refractive index deviation;

[0088] The air refractive index deviation is input to the displacement compensation unit, which maps the refractive index deviation into a displacement correction amount based on a preset calibration coefficient.

[0089] Based on the purification displacement value, a displacement correction value is subtracted from the purification displacement value to finally generate the corrected displacement value. Specifically, the system first collects environmental data in real time using air pressure, temperature, and humidity sensors installed at the measurement site, and uses low-pass filtering to remove high-frequency noise from the original signal, forming environmental state data with timestamps. Subsequently, the system inputs this data into a polynomial approximate refractive index model established during the calibration phase. This model is offline fitted based on a large amount of experimental data using international standard formulas and can output the current air refractive index in real time on an embedded platform using efficient polynomial calculation. The system calls the reference air refractive index recorded during the factory or periodic calibration phase and compares it with the real-time calculation result to obtain the air refractive index deviation. This deviation enters the displacement compensation unit, which converts the refractive index deviation into a corresponding displacement correction amount based on the linear mapping coefficients obtained during calibration, and performs necessary smoothing processing on the correction amount to eliminate instantaneous fluctuations. Finally, the system subtracts this displacement correction amount from the purification displacement value output by S4 to generate the final corrected displacement value, and adds a timestamp, current environmental parameters, refractive index deviation, and correction amount metadata to this value. The correction displacement value is transmitted to the host computer or control system in real time for high-precision machining or measurement tasks, thereby ensuring that laser interferometric displacement measurement maintains metrological-grade accuracy and stability in variable environments.

[0090] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A high-precision method for detecting the end-effector displacement of an industrial robotic arm based on photoelectric interferometry, characterized in that, include: S1: Acquire the two orthogonal interference electrical signals output by the photoelectric interference module, perform digital down-conversion and filtering processing, and obtain the complex envelope signal; S2: Based on the complex envelope signal, differential operations are performed to obtain the instantaneous phase increment, which is then accumulated to obtain the accumulated optical phase. According to the mapping relationship between the accumulated optical phase and the laser wavelength, the preliminary displacement observation value is calculated, and the displacement observation base set is constructed. The acquisition of the accumulated optical phase includes: firstly, performing short-time complex domain smoothing on the complex envelope signal sequence to suppress high-frequency noise, and then calculating the difference between adjacent sampling points to construct a complex difference; by obtaining the imaginary part of the ratio between the complex difference and the complex envelope signal, the instantaneous phase increment is obtained based on the first-order approximation principle; the instantaneous phase increment is accumulated over time to form the accumulated optical phase based on the initialization time. The initial value of the complex envelope signal is set by the system sampling under static conditions, and the accumulated optical phase is set to zero during phase accumulation initialization. S3: Collect equipment information to construct a sensor dataset and generate a multi-source error state vector; S3 includes: collecting multi-source sensor data from the equipment to construct a device sensor dataset, which includes: joint position number, joint position coordinates, joint angle, joint and structure temperature, triaxial accelerometer signal, and gyroscope signal; calculating the theoretical end position based on the joint position number, joint position coordinates, joint angle, preliminary displacement observation, and preset nominal kinematic parameters; differentiating the theoretical end position from the preliminary displacement observation to obtain a residual signal; estimating the correction amount related to the nominal kinematic parameters based on the residual signal; mapping the correction amount to kinematic deviation; and calculating the temperature of each joint and structure of the robotic arm and the operating time of the actuators. A thermal state data sequence is constructed. Based on the thermal state data sequence and a preset thermal expansion model, the displacement change caused by structural thermal expansion is calculated, and the displacement change is set as the thermally induced displacement deviation. Based on the triaxial accelerometer and gyroscope signals at the end of the robotic arm, a vibration state data sequence is constructed. Based on the vibration state data sequence and the structural dynamics model, the deviation of the end-effector vibration position is calculated, and the deviation is set as the vibration displacement deviation. Based on the time series of the vibration displacement deviation at the end of the robotic arm, the vibration velocity change at continuous moments is calculated, or the vibration velocity information is directly extracted using the gyroscope signal, and the calculation result is used as the vibration velocity term. A multi-source error state vector is generated by combining the kinematic deviation, thermally induced displacement deviation, vibration displacement deviation, and vibration velocity term. S4: Based on the extended Kalman filter algorithm, the multi-source error state vector is optimally estimated, and the optimally estimated multi-source error state vector is separated from the preliminary displacement observations to generate the purified displacement value; S5: Calculate the air refractive index based on the real-time collected environmental parameters, and compensate the purification displacement value by combining it with the preset reference air refractive index, and finally generate the corrected displacement value.

2. The high-precision detection method for the end-effector displacement of an industrial robotic arm based on photoelectric interferometry according to claim 1, characterized in that, S1 includes: A photoelectric interference module is installed at the end of the robotic arm to emit a measurement beam that is directed at a fixed reference mirror. The measurement beam is a laser with a fixed wavelength. The interference light signal is received by a photodetector and converted into two orthogonal interference voltage signals. The two synchronously sampled orthogonal interference voltage signals are then combined into a complex measurement signal. The complex measurement signal is multiplied by a preset digital local oscillator sequence to perform digital down-conversion, wherein the angular frequency of the digital local oscillator sequence is equal to the frequency shift angular frequency of the acousto-optic modulator in the photoelectric interference module; The down-converted signal is subjected to digital low-pass filtering to obtain a complex envelope signal.

3. The high-precision detection method for the end-effector displacement of an industrial robotic arm based on photoelectric interferometry according to claim 1, characterized in that, Real-time monitoring of complex envelope signals, and extraction of the magnitude and amplitude of complex envelope signals as quality indicators; When the quality index is lower than the preset threshold, the signal quality is determined to be faulty, the phase accumulation process is paused and the resampling process is triggered.

4. The high-precision detection method for the end-effector displacement of an industrial robotic arm based on photoelectric interferometry according to claim 3, characterized in that, When the quality index is not lower than the preset threshold, the cumulative optical phase is multiplied by the preset proportional coefficient to obtain the preliminary displacement observation value. The proportional coefficient is the ratio of the laser vacuum wavelength to a constant four times pi. The laser vacuum wavelength is taken from the laser parameters of the photoelectric interference module. Add the current system timestamp and quality index to each preliminary displacement observation; By integrating accumulated optical phase, preliminary displacement observations, timestamps, and quality indicators, a basic set of displacement observations is constructed.

5. The high-precision detection method for the end-effector displacement of an industrial robotic arm based on photoelectric interferometry according to claim 1, characterized in that, S4 includes: A state transition model and an observation model are established based on the multi-source error state vector, and the extended Kalman filter method is used to estimate and update the multi-source error state vector in real time. The Kalman-filtered and updated multi-source error state vector is removed from the initial displacement observations to obtain the purified displacement value.

6. The high-precision detection method for the end-effector displacement of an industrial robotic arm based on photoelectric interferometry according to claim 1, characterized in that, S5 includes: Real-time environmental parameters are collected by barometric pressure sensors, temperature sensors, and humidity sensors to construct environmental status data; Environmental state data is input into a preset refractive index estimation model, which outputs the current air refractive index. The refractive index estimation model is constructed based on a polynomial approximation method.

7. The high-precision detection method for the end-effector displacement of an industrial robotic arm based on photoelectric interferometry according to claim 6, characterized in that, The reference air refractive index preset during the system calibration phase is called, and the current air refractive index is compared with the reference air refractive index to obtain the air refractive index deviation; The air refractive index deviation is input to the displacement compensation unit, which maps the refractive index deviation into a displacement correction amount based on a preset calibration coefficient. Based on the purification displacement value, the displacement correction value is subtracted from the purification displacement value to finally generate the correction displacement value.

Images